Vision-Based Door Operation for Human Intent Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contemporary sensor-based automatic door operating systems frequently generate false openings, leading to energy waste and reduced equipment lifetime due to inaccurate detection of objects, such as animals or passing pedestrians.
Innovation Solution
A door operating system that utilizes a control unit and optical sensors, like cameras, to process observation image data for detecting individuals, extracting pose and viewing direction information, and determining the probability of intent to pass the door, operating the door only when the probability exceeds a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If sensor-based automatic door operating systems use infrared, ultrasonic/radio or other wireless sensing methods to detect objects, then the door can be operated automatically, but false opening operations occur frequently due to inaccurate detection of objects such as animals or passing pedestrians
Solution Approach 1:
The system segments the detection process into multiple independent analysis stages: object detection, pose estimation, viewing direction analysis, and probability calculation. Each stage processes specific features and contributes to the final decision, allowing the system to distinguish between genuine users and false triggers by evaluating multiple characteristics simultaneously
Solution Approach 2:
The system transitions from traditional single-dimension detection (presence/absence) to multi-dimensional analysis by incorporating pose information, viewing direction, and probability values. This dimensional expansion enables more nuanced discrimination between actual users and false detections, significantly improving detection accuracy while maintaining automation
2Extent of automation
If the door operates frequently based on sensor detection, then automatic operation is achieved, but energy waste occurs particularly in air conditioning energy
Solution Approach 1:
The system performs preliminary analysis of pose and viewing direction information before triggering door operation. By evaluating the probability value in advance based on multiple characteristics, the system prevents unnecessary door openings caused by false detections, thereby avoiding associated energy waste from air conditioning and other building systems
Solution Approach 2:
The system incorporates feedback mechanisms where the control unit continuously evaluates detection data, pose information, and viewing direction to calculate probability values. This feedback loop enables the system to learn from patterns and make more accurate decisions, reducing false operations and associated energy consumption over time
3Extent of automation
If the door operates frequently based on sensor detection, then automatic operation is achieved, but equipment lifetime is reduced
Solution Approach 1:
The system performs preliminary evaluation of detection accuracy using pose and viewing direction analysis before executing door operation. This pre-assessment prevents unnecessary operations triggered by false detections, thereby reducing wear and tear on the door mechanism and extending equipment lifetime
Solution Approach 2:
The system uses disposable or easily replaceable sensor components for initial detection, while relying on more durable control unit processing for final decision-making. This strategy protects the main door operating mechanism from excessive use by filtering out false detections at the processing stage
Data Source
AI summary
A door operating system including a control unit and at least one sensor device, the system configured to operate a door from a closed state into an opened state. The sensor device is configured to observe an observation area and provide image data to the control unit, which is configured to receive and process the observation image data. Processing the data includes detecting an individual inside the observation area, extracting pose information and/or viewing direction information corresponding to the individual, determining, based on the pose information and/or the viewing direction information, a probability value representing the probability that the individual intends to pass the door, and determining whether the probability value is higher than a probability threshold value, wherein the door operating system operates the door from a closed state in an opened state if the control unit determines that the probability value is higher than a probability threshold value.